Open Access
Add to BibSonomyAbstract
We show here that live e-coli bacterial culture, thanks to the self-propelling feature, can significantly reduce the coherent noise. In fact, the typical self-propelled drive of such microorganisms provides enough time diversity in speckle patterns. Optical properties of a bacteria suspension have been investigated and analyzed thus showing that it behaves as a quite good optical speckle decorrelation device. Samples with different bacteria densities have been studied. The decorrelation effect has been demonstrated by probing the imaging performance in through transmission in coherent microscope configuration.
© 2015 Optical Society of America
1. Introduction
A live bacteria suspension is considered here as a sort of turbid medium in which the self-propelling feature of its individual elements causes a whole time diversity in speckle patterns. Speckle decorrelators are widely employed to reduce the coherence and improve the quality of imaging systems and optical projectors [1,2]. Usually, moving diffusers placed along the optical beam path are used for the scope [1]. Differently, the source can be fruitfully engineered exploiting colloidal suspensions to obtain random lasers with significantly reduced speckle [2]. On the other hand, imaging through turbid media in Lab-on-Chip (LoC) devices [3–5] has been deeply investigated in recent papers and different solutions have been proposed in some practical cases for restoring clear imaging [6–12]. In most cases, the turbidity is due to the presence of inanimate micro or sub-micro sized objects embedded into the fluid, e.g. colloidal particles, responsible for severe light scattering. However, sometimes, the turbid volume can be constituted of living material, such as bacteria colonies, which could grow in favorable environmental conditions, e.g. during the monitoring of slow biological processes [13]. In the case of stationary layers, the scattering problem can be tackled whenever the transmission matrix of the layer can be characterized [9,10]. However, in the case of turbid liquids, even if quasi-static, the transmission matrix is not available and a different strategy has to be adopted. For example, the Brownian motion of suspended sub-micro objects can be exploited to acquire enough uncorrelated speckle-like patterns and clear coherent imaging is achievable by multi-look Digital Holography (DH) [6–15]. Unfortunately, for big size scattering objects, the Brownian motion does not provide enough time-diversity and such strategy fails [7]. Hence, the holographic capability of seeing through turbid volumes was limited to flowing objects or small randomly moving particles. However, in case of turbid volume constituted of living material, a different case can be considered, whenever the living objects exhibits a self-propelling feature. In particular, here we study the interesting situation in which live bacteria are responsible of light scattering. Indeed, their dynamic behavior obeys different displacement laws if compared to the inanimate micro-particles and thus a different speckles diversity effect is expected. Investigation on the interaction of light with bacteria has been carried out by means of scattering and diffraction characterizations [16–22]. Scattering properties of bacteria are of great importance and have been intensively studied in the field of environmental monitoring for early detection of biological contamination with the aim to prevent the spread of diseases, or even in cutting-edge applications as photobioreactors to constitute elements of photosynthetic-plasmonic voltaic cells [16,17]. Moreover, great effort has been spent to find reliable ways to distinguish between different bacteria species. In [18] estimates of diameter and length of single E-coli have been performed by Inverse Light Scattering (ILS), relying on prior assumptions about the characteristic shape of the investigated bacteria culture. Elastic Light Scattering (ELS) has been proved to accurately estimate bacteria size by multiple measures at different angles [17]. Noteworthy, a reliable classification of different species has been achieved by statistical analysis of the diffraction pattern produced by bacteria biofilms [19,20]. However, the research about thick and severely scattering live bacteria volumes in terms of their scattering properties and how these affect the imaging, i.e. by means of an ensemble characterization of the speckle pattern, has never been considered before. Bacteria contamination of a liquid volume results in severe scattering and strongly hinders the capabilities of any imaging system. Actually, the characteristic size of the scattering objects considered in our case ranges from 1 μm to 3 μm (thickness), depending on the considered species. Moreover, bacteria typically organize in longer aggregates (from 10 μm to 100μm), as shown in the microscope image of Fig. 1, so that the uncorrelation due to the Brownian motion cannot be exploited. Nevertheless, bacteria have the unique feature to be self-propelling microorganisms. When observed at the microscope, they show frenetic and random movements (see Media 1) that are expected to be uncorrelated at long time scales. In fact, we show that the remarkable decorrelation effect of live microorganisms can be exploited, to accomplish such an optical task. As a test bed, experimental demonstration of the speckle diversity is performed through coherent holographic microscopy, showing that clear imaging through a bacteria suspension is actually achievable along with the holographic flexible numerical focusing capability. Thus, this work unlocks a new imaging option to study bio-samples, which could not be observed through standard optical microscopy, whenever they are immersed in a liquid volume contaminated by bacteria.
Fig. 1 (Media 1) DH set-up employed to image a test target through a severely scattering bacteria volume. BS: Beam Splitter. M: Mirror. BE: Beam Expander. BC: Beam Combiner. MO: Microscope Objective. S: Sample. On the bottom-right corner a 20x microscope image of the bacteria suspension is shown.
At first, we analyze the time correlation properties of the bacteria culture showing different statistics with respect to the case of RBCs (i.e. occluding objects with comparable characteristic length) [7]. Starting from this characterization, we demonstrate that the speckle decorrelation effect, due to the self-propelling feature of the elements in the suspension, is sufficient to restore clear imaging, allowing us to take advantage of a multi-look strategy without the need for prior information or scattering measures [9,10]. In this way we can afford a challenging quasi-static case, where the typical dimensions of the occluding objects would impair any clear imaging as the Brownian motion is not of help.
2. Working principle
The experimental set-up is shown in Fig. 1. This is a Mach-Zehnder interferometer in transmission configuration. The coherent laser source emits a beam at 532nm wavelength, which is split in two parts, namely the reference beam and the object beam. The object beam goes through the sample before being recombined with the reference in order to form an interference pattern in the acquisition plane, which is collected by a CCD camera with 6,45μm pixel size. In our experiments, an USAF test resolution target has been placed behind a Petri dish filled with a bacteria culture, as shown in Fig. 1 (details about the sample preparation are provided in the next Section). Let and be the reference beam and the object beam in the acquisition plane The digital hologram recorded at time t can be expressed as
where each of the last two terms carries the whole complex information of the object and can be extracted by spatial filtering in the Fourier domain. In Eq. (1) the spatial variables have been omitted for the sake of brevity, while the time dependence takes into account the ensemble effect of the bacteria volume which is a degradation of the hologram as a time variable speckle pattern superposing to the useful object information. Hence, a characterization is required to study the time correlation of the speckle patterns. In particular, we measured the average correlation coefficient defined as:i.e. the average of the correlation coefficients measured between all the holograms couples acquired at time lags τ. In Eq. (2) N is the number of looks, denotes the spatial expectation operator adopted to calculate each correlation coefficient, and i = rt where r is the frame rate. In Fig. 2 the trend of vs. is shown, compared with the correlation coefficient obtained in the case of channel filled with quasi-static RBCs [7]. A saturation value is reached (see the blue solid line), while this keeps close to the 90% for the case of the RBCs (red solid line). This result suggests us to exploit the self-propelling property of bacteria to acquire and incoherently combine a stack of uncorrelated hologram reconstructions. Once the hologram stack is acquired, numerical propagation provides theSingle-Look (SL) stack of images in the object focus plane and the Multi-Look (ML) amplitude reconstruction, defined aswhere denotes the propagation transform (in DH processing the preferred reconstruction strategies are based on the Fresnel transform or the Angular Spectrum method each one aimed at solving numerically the Fresnel-Kirchhoff diffraction integral [15]). Moreover, a ML quantitative phase-contrast map can be retrieved from the same stack as in [8] whenever pure-phase biological specimens are objects of study. Thanks to the possibility to obtain a ML image which is less affected by noise due to the bacteria scattering, and taking advantage of the DH flexible focusing capability, the proposed approach also allows an automatic numerical focusing to look for the proper object best-focus plane. In fact, an autofocusing algorithm [23] can be simply applied to ML images obtained by the incoherent combination of holograms propagated at different distances. Indeed, the best-focus plane can be obtained optimizing the Tamura coefficient, [23]. This is known to be a proper contrast estimator for the scope as is a function of z exhibiting a maximum at the object best-focus plane. In particular, the focus distance can be estimated aswhere z is the axial distance, is the stack of ML amplitude reconstructions propagated at various distances after selecting the diffraction order of interest. In the next Section we show and discuss the results of experiments carried out with the aim to demonstrate the possibility to achieve clear imaging with flexible autofocusing capabilities, starting from blind out-of focus recordings of targets hidden by a severely scattering bacteria suspension.3. Experiments
E. coli DH5-alpha was plated and incubated on agar plates. The day before the beginning of experiment, a single bacterial colony was picked and cultured in LB broth medium at 37°C in a shaker incubator for 16-18 hours to achieve saturation conditions. A 1:5 volumetric dilution of cell culture was then grown in LB until reaching the log phase. Then the growth was stopped and bacteria were harvested by centrifugation at 5000 rpm for 10 min in order to separate the cells from the medium. Cells were then resuspended in fresh LB in order to obtain a final concentration of 6^106 cells/mL. Bacteria concentration was calculated by spectrophotometric measurement of suspension absorbance at 600nm (Optical Density at 600nm, i.e. OD600), considering that 8*108 cell/mL have an OD600 = 1.
In our tests, N = 500 holograms have been registered with frame rate r = 12s−1 and the experiment has been repeated progressively decreasing the dilution percentage, i.e. the density of bacteria present inside the analyzed suspension. In the beginning, the ensemble effect of a thin bacteria layer (200μm thickness) has been investigated studying the time variation of the hologram spectra. The test object and the corresponding hologram are shown in Fig. 3(a). From the digital hologram the speckle-like pattern superposing to the useful signal is apparent. Noteworthy, the analysis of the hologram spectra (Media 2) clearly shows the ensemble effect of the scattering suspension, as higher spatial frequencies are visible constituting a cloud superposing to the useful signal. Unfortunately, the scattering noise overlaps to the higher object frequencies in the hologram spectrum, as shown in Fig. 3(b), so that this cannot be simply discarded by band-pass filtering in the Fourier domain. However, the time-analysis of the spectra confirms the indications of the correlation coefficient trend, revealing the variability of the spurious contributions and suggesting that a ML gain can effectively enhance the useful signal. Figure 3(c) shows the ML amplitude reconstruction in the object focus plane, where, as expected, the hidden target is well visible. In a second experiment we imaged a target (200 lines/mm) through volumes of liquid filled with bacteria (a 500μm thick container was used for the scope) at different concentrations, ranging from 1.4x105 cell/mL to 2x105 cell/mL. Figure 4 shows the corresponding SL and ML amplitude reconstructions. In particular, the SL images look severely degraded even in case of low bacteria concentration (dilution higher than 80% in Figs. 4(a) and 4(d), corresponding to 1.4x105 cell/mL and 1.5x105 cell/mL respectively) while a high density bacteria suspension (2x105 cell/mL) completely hides the target, as shown in Fig. 4(g). On the contrary, the ML strategy is successful in restoring the useful object signal, providing enhanced quality images (in terms of noise rejection) in case of 1.4x105 cell/mL [see Fig. 4(b)] and 1.5x105 cell/mL [Fig. 4(e)], and clearly revealing the hidden target in case of volumes with high bacteria density (i.e. 2x105 cell/mL), as shown in Fig. 4(h).
Fig. 3 Imaging a test target through a thin scattering bacteria layer. (a) Hologram of the target shown in the inset. (b) (Media 2) Hologram spectrum ( + 1 order). (c) ML reconstruction.
Fig. 4 (Media 3) Amplitude reconstructions of a test target (200 lines/mm) through bacteria volumes at different concentrations. (a,d,g) SL images. (b,e,h) ML images. (c,f,i) Image contrast over the lines indicated with the horizontal bars. Blue: SL image. Red: ML image.
This is also apparent from the plots reported in Figs. 4(c), 4(f) and 4(i) showing the image amplitude signals, measured over the horizontal bars indicated in the corresponding SL and ML images. The blue lines refer to the SL, while the red lines correspond to the ML images. In particular, in Figs. 4(c) and 4(f) the ML gain results in a contrast enhancement on the vertical lines and, hence, an improved resolution. Moreover, the structure of the lines becomes resolvable in the ML plot of Fig. 4(i), while these are not appreciable at all in the SL plot, as the useful signal is entirely covered by noise. Obviously, the boundary dilution is set by the power of the beam reaching the biological samples and the target. If the laser power is increased, clear imaging is expected although the boundary condition is overcome, as the detector collects unscattered photons with higher probability. As a performance measure, we calculated the Dispersion Index (DI) [7,8], over an homogeneous segment of the image [red box in Fig. 4(g)], so that a dispersion decrease is expected as a result of noise mitigation. We found a remarkable ML improvement with respect to the SL amplitude (up to 75% decrease of DI, as shown in Fig. 5(a)). Media 3 shows the ML improvement we achieve when increasing the number of looks, i.e. the number of holograms we combine to synthesize the ML reconstruction, in the case of the acquisitions of Figs. 4(g) and 4(h). Some of the most significant ML reconstructions are also reported in Fig. 5(a). Noteworthy, the DI saturation shows that a lower number of acquisitions is sufficient to provide the desired image enhancement, thus allowing us to save computational time. If a proper decimation of the hologram stack is performed, this time can be further reduced without losing ML gain. In order to show the flexible focusing capability of MLDH, two further experiments have been carried out. In particular, we inserted a hair along the object beam path at very long distance from the target. Starting from the same stack of holograms, propagations at different distances provide ML reconstructions where the target [Fig. 5(c), Media 4] or the hair [Fig. 5(d)] are clearly imaged in their own best focus planes, while these objects are not visible in the SL reconstructions of Fig. 5(b). Such capability allows one to acquire holograms of objects placed in different positions along the optical axis, and to reconstruct them in a condition where it is impossible to determine each object focus plane by mechanical scanning. With the aim to demonstrate the full automatic refocusing of the object, we applied the contrast optimization expressed in Eq. (4) to the acquisitions corresponding to Fig. 4(d) (concentration of 1.5x105 cell/mL). In Fig. 6 the plot of vs. is reported, along with some ML reconstructions after propagation at various distances (see the insets in Fig. 6), showing that the z maximizing the contrast actually corresponds to a ML reconstruction where the test object is in focus. On the contrary, in the SL reconstruction shown in the bottom left corner of Fig. 6 the object signature is just slightly recognizable, impairing the convergence of any autofocusing algorithm. Hence, it becomes possible to fully exploit the flexible holographic refocusing to obtain images of hidden objects in focus, in a condition where it would not be possible by conventional microscopy. Indeed, this would require a mechanical scanning along the optical axis to look for , but the presence of the scattering volume would impair the formation of a recognizable image of the object in all the inspected planes.
Fig. 5 (Media 3) (a) Dispersion index [%] vs. the number of looks, N. In the insets some of the corresponding N-looks amplitude reconstructions are shown. (b-d) To show MLDH flexible focusing a hair is inserted along the object beam path. (b) SL, target on focus. (c) (Media 4) ML, target on focus. (d) ML, hair on focus.
Fig. 6 Tamura coefficient vs. z[cm]. The inset on the bottom-left corner shows the SL reconstruction. The insets indicated by the vertical arrows show ML reconstruction after propagation at various distances.
4. Conclusions
In this work, we investigated how a liquid volume contaminated by bacteria affects the capabilities of a coherent imaging system. Stationary bacteria produce harsh scattering conditions and, in the most of the cases, impairs any monitoring of the processes happening in LoC platforms. Nevertheless, differently from the case of static colloidal suspensions made of inanimate objects, the self-propelling properties of live bacteria gives one the unique chance to recover the whole complex information of hidden objects, as the movement of such microorganisms provides enough decorrelation between the acquired speckle patterns. For this reason, such live objects are able to accomplish an optical task, i.e. they allow to compensate the optical wavefront distortion caused by themselves. Experiments have been carried out imaging test targets placed behind bacteria at different concentrations, showing that a multi-look DH strategy can be successfully adopted for the scope. Starting from blind out-of focus recordings, automatic refocusing through scattering bacteria has been achieved without the need of a mechanical z-scanning, which can be exploited to look for hidden biological objects whose presence and/or position along the optical axis is not known a priori. Noteworthy, we found that the presence of bacteria suspension on the analyzed sample allows to lower the coherence of the imaging systems, enhancing the image quality in terms of signal to noise ratio [8].
Acknowledgments
This work was supported by Progetto “Smart Cities” – AquaSystem: PON04a2
References and links
1. F. Monroy, O. Rincon, Y. M. Torres, and J. Garcia-Sucerquia, “Quantitative assessment of lateral resolution improvement in digital holography,” Opt. Commun. 281(13), 3454–3460 (2008). [CrossRef]
2. B. Redding, G. Allen, E. Dufresne, and H. Cao, “Low-loss high-speed speckle reduction using a colloidal dispersion,” Proc. Of CLEO: Science and Innovations, San Jose, California (US), 2013.
3. A. Ozcan and U. Demirci, “Ultra wide-field lens-free monitoring of cells on-chip,” Lab Chip 8(1), 98–106 (2007). [CrossRef] [PubMed]
4. X. Heng, D. Erickson, L. R. Baugh, Z. Yaqoob, P. W. Sternberg, D. Psaltis, and C. Yang, “Optofluidic microscopy--a method for implementing a high resolution optical microscope on a chip,” Lab Chip 6(10), 1274–1276 (2006). [CrossRef] [PubMed]
5. L. Schmid, D. A. Weitz, and T. Franke, “Sorting drops and cells with acoustics: acoustic microfluidic fluorescence-activated cell sorter,” Lab Chip 14(19), 3710–3718 (2014), doi:. [CrossRef] [PubMed]
6. V. Bianco, M. Paturzo, A. Finizio, D. Balduzzi, R. Puglisi, A. Galli, and P. Ferraro, “Clear coherent imaging in turbid microfluidics by multiple holographic acquisitions,” Opt. Lett. 37(20), 4212–4214 (2012). [CrossRef] [PubMed]
7. V. Bianco, M. Paturzo, A. Finizio, A. Calabuig, B. Javidi, and P. Ferraro, “Clear microfluidics imaging through flowing blood by digital holography,” IEEE J. Sel. Top. Quantum Electron. 20(3), 1–7 (2013).
8. V. Bianco, F. Merola, L. Miccio, P. Memmolo, O. Gennari, M. Paturzo, P. A. Netti, and P. Ferraro, “Imaging adherent cells in the microfluidic channel hidden by flowing RBCs as occluding objects by a holographic method,” Lab Chip 14(14), 2499–2504 (2014). [CrossRef] [PubMed]
9. V. Bianco, M. Paturzo, O. Gennari, A. Finizio, and P. Ferraro, “Imaging through scattering microfluidic channels by digital holography for information recovery in lab on chip,” Opt. Express 21(20), 23985–23996 (2013). [CrossRef] [PubMed]
10. Y. Choi, C. Yoon, M. Kim, W. Choi, and W. Choi, “Optical imaging with the use of a scattering lens,” IEEE J. Sel. Top. Quantum Electron. 20(2), 6800213 (2013).
11. J. Bertolotti, E. G. van Putten, C. Blum, A. Lagendijk, W. L. Vos, and A. P. Mosk, “Non-invasive imaging through opaque scattering layers,” Nature 491(7423), 232–234 (2012). [CrossRef] [PubMed]
12. X. Yang, Y. Pu, and D. Psaltis, “Imaging blood cells through scattering biological tissue using speckle scanning microscopy,” Opt. Express 22(3), 3405–3413 (2014). [CrossRef] [PubMed]
13. A. Khmaladze, R. L. Matz, T. Epstein, J. Jasensky, M. M. Banaszak Holl, and Z. Chen, “Cell volume changes during apoptosis monitored in real time using digital holographic microscopy,” J. Struct. Biol. 178(3), 270–278 (2012). [CrossRef] [PubMed]
14. Y. Cotte, F. Toy, P. Jourdain, N. Pavillon, D. Boss, P. Magistretti, P. Marquet, and C. Depeursinge, “Marker-free phase nanoscopy,” Nature Photon. 7(2), 113–117 (2013). [CrossRef]
15. P. Memmolo, V. Bianco, M. Paturzo, B. Javidi, P. A. Netti, and P. Ferraro, “Encoding multiple holograms for speckle-noise reduction in optical display,” Opt. Express 22(21), 25768–25775 (2014). [PubMed]
16. N. Samsonoff, M. D. Ooms, and D. Sinton, “A photosynthetic-plasmonic-voltaic cell: excitation of photosynthetic bacteria and current collection through a plasmonic substrate,” Appl. Phys. Lett. 104(4), 043704 (2014). [CrossRef]
17. A. Katz, A. Alimova, M. Xu, E. Rudolph, M. K. Shah, H. E. Savage, R. B. Rosen, S. A. McCormick, and R. R. Alfano, “Bacteria size determination by elastic light scattering,” IEEE J. Sel. Top. Quantum Electron. 9(2), 277–287 (2003). [CrossRef]
18. A. I. Konokhova, A. A. Gelash, M. A. Yurkin, A. V. Chernyshev, and V. P. Maltsev, “High-precision characterization of Individual E. coli cell morphology by scanning flow cytometry,” Cytometry A 83(6), 568–575 (2013). [CrossRef] [PubMed]
19. A. Suchwalko, I. Buzalewicz, and H. Podbielska, “Computer-based classification of bacteria species by analysis of their colonies Fresnel diffraction patterns,” Proc. Of SPIE,” Frontiers in Biological Detection: From Nanosensors to Systems IV. 8212, 82120R1-82120R13 (2012).
20. E. Ponder, “Diffraction pattern produced by bacteria,” J. Exp. Biol. 11, 54–57 (1933).
21. B. Rajwa, M. M. Dundar, F. Akova, A. Bettasso, V. Patsekin, E. D. Hirleman, A. K. Bhunia, and J. P. Robinson, “Discovering the unknown: detection of emerging pathogens using a label-free light-scattering system,” Cytometry A 77(12), 1103–1112 (2010). [CrossRef] [PubMed]
22. J. Miao, K. O. Hodgson, T. Ishikawa, C. A. Larabell, M. A. LeGros, and Y. Nishino, “Imaging whole Escherichia coli bacteria by using single-particle x-ray diffraction,” Proc. Natl. Acad. Sci. U.S.A. 100(1), 110–112 (2003). [CrossRef] [PubMed]
23. P. Memmolo, M. Paturzo, B. Javidi, P. A. Netti, and P. Ferraro, “Refocusing criterion via sparsity measurements in digital holography,” Opt. Lett. 39(16), 4719–4722 (2014). [PubMed]
